Professional Documents
Culture Documents
ABSTRACT The integration of non-synchronous generation units and energy storage through power
electronics is introducing new challenges in power system dynamics. Specifically, the rotor angle stability
has been identified as one of the major obstacle with regards to power electronics dominated power systems.
To date, conventional power system stabilizer (PSS) devices are used for damping electromechanical
oscillations, which are only tuned sporadically leading to significant deterioration in performance against
the ever-changing operating conditions. In this paper, an intelligent power oscillation damper (iPOD) is
proposed for grid-forming converters to attenuate electromechanical inter-area power oscillation. In partic-
ular, the iPOD is applied to a synchronous power controller (SPC) based grid-forming power converter to
increases gain of the active power control loop at the oscillatory frequency. Predictions regarding the mode
frequency, corresponding to the current operating points, are given by an artificial intelligence ensemble
model called Random Forests. The performance of the proposed controller is verified using the two area
system considering symmetrical fault for random operating points. In addition, a comparison with PSS
installed in each generator reveals the individual contribution with respect to the inter-area mode damping.
INDEX TERMS Artificial intelligence, ensemble modelling, inter-area power oscillation, random forests,
synchronous power controller.
I. INTRODUCTION Conventionally, the PSS has been used to damp low fre-
Renewable-based Energy Systems (RES) and Energy Storage quency oscillations in a system by regulating the synchronous
Systems (ESS) are widely adopted to reduce green house generator’s excitation voltage.Among the standardized PSS
gas emissions and improve efficiency. These systems, due to schemes, the classical PSS1 controller consists of three main
their incompatibility with synchronous frequency (50/60 Hz), blocks: gain, washout filter and phase compensation. Consid-
require power electronic devices to act as an intermedi- ering inter-area modes, the time constant of the washout filter
ate before connection with the rest of the network is pos- needs to be carefully chosen, since it captures the low fre-
sible. Consequently, as the presence of power electronics quency oscillations [3]. The PSS1 uses a single input signal,
increases, the power system dynamics are affected intro- typically the speed deviation 1ω. More sophisticated designs
ducing new challenges regarding power system stability [1]. are the PSS2B and PSS4B controllers. In PSS2B, speed
In fact, ENTSO-E has identified the angular stability of transducers are added to improve reactive power modulation
power systems as one of the key challenges related to High during mechanical power changes and eliminate torsional
Penetration of Power Electronic Interfaced Power Sources forces that can harm the generator shaft. PSS4B builds on
(HPoPEIPS) [2]. top of the PSS2B by incorporating low, intermediate and high
frequency bands to broaden the range of targeted modes [4].
The associate editor coordinating the review of this manuscript and One major limitation of the PSS is the high number of
approving it for publication was Xiaorong Xie . parameters that need to be tuned for attenuating system
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
VOLUME 8, 2020 93369
G. N. Baltas et al.: Grid-Forming Power Converters Tuned Through Artificial Intelligence
oscillations [5]. For instance, a typical PSS has 8 parameters: components of the targeted mode of oscillation as it moves
gain, time constant of wash out filter, and four constants for according to the current operating conditions. The main con-
lead-lag compensator. These parameters usually are tuned at tribution of the system, presented in the following, is a novel
commissioning of the power plant and remain fixed despite control scheme that incorporates intelligence by exploiting
the changes of operating conditions [6]. There is an open the ability of AI to predict the characteristic values of an
discussion in the literature on possible methods for adaptive oscillatory mode in real time and uses this information to
PSS and optimal tuning [3]. This becomes even more relevant adaptively tune a dedicated loop of the SPC to damp oscilla-
now, as power electronics can alter the oscillatory profile of tions and enhance the system stability i.e. iPOD. In this work,
a given system, changing or creating new modes. special focus is put on describing all the steps for developing
Currently, the number of synchronous generators is slowly the AI models in detail. The proposed approach is verified
phased out [7] and as a result so will the number of PSS. Yet, under a symmetrical fault for random operating points as
power electronic interfaced generation units can provide a well as against conventional PSS devices using the well-
fast response to the grid and assist the system in maintaining known two area system described in [21]. The AI predictor
its stability in case of a disturbance [8]. Such grid friendly is trained for tracking the inter-area mode between the two
response can be produced by means of the currently well- areas, however the same principles can be applied to develop
known grid-forming power converters. In general, the control multiple AI predictors to predict more than one inter-area
system of a grid-forming converter can either be based on modes.
droop [9] or on virtual synchronous machine (VSM) [10]. The remainder of this paper is structured as follows.
In the former case, the contribution of the power converter is Section II provides a detail description regarding the test case.
usually limited to primary frequency and voltage regulation. Section III presents the proposed AI-based adaptive SPC
To add grid-supporting functionalities, e.g. inertia emulation model. Section IV analyse the results and Section V discusses
and power oscillation damping, the VSM implementation can the conclusions.
be adopted in the grid-forming power converter [11].
Among the various VSM strategies, reported in the lit- II. TEST CASE
erature, the Synchronous Power Controller (SPC) has been A popular power system to analyse the behavior of an inter-
widely used mainly due its simple yet effective control area mode is the two-area system [21], depicted in Fig. 1.
structure [12]. By employing the SPC, power converters can The system consists of two areas with one load (L07, L09),
provide virtual damping and synthetic inertia to the grid, one shunt capacitor (C7, C9) and two generators (G1-G2,
increasing its dynamic flexibility to improve local and inter- G3-G4), respectively. Each generator is equipped with IEEE
area stability [13], [14]. In fact, the SPC with multiple power Type AC4A Excitation System (AVR), steam turbine gov-
loop controllers (PLC) has been presented in [15]. By com- ernors (GOV) and PSS type 1. These areas are connected
bining a band-pass filter with a modified swing equation, through a weak tie-line where 400MW are flowing from Area
the inter-area oscillations at a predefined frequency can be 1 to Area 2. Modal analysis reveals two local modes with
attenuated. However, to ensure the system in [15] works frequencies of 1.05Hz and 1.08Hz in Area 1 and Area2,
effectively, the center frequency of its band-pass filter has respectively, as well as an inter-area mode with frequency of
to be tuned properly to match the frequency of the inter-area approximately 0.55Hz.
oscillation mode, which is usually not accessible due to the The specific design of this system not only allows for
dynamic behavior of the power system. studying the behaviour of electromechanical oscillations
Recently, data-driven methods based on artificial intelli- between two areas, but also on demonstrating the effective-
gence (AI) are studied for accurately and efficiently deter- ness of control methods (such as the iPOD) for attenuating
mining the state of the power system in real time and thereby them. To demonstrate the impact of the iPOD, a virtual-
avoid instability [16]. Specifically, the relationship between synchronous generation unit equiped with an SPC is con-
system variables and stability indexes can be modeled to pre- nected at bus 7, through two step-up transformers. The rated
dict the post-fault state of the system. For instance, an ensem- apparent power of the aforementioned unit is 100MVA, equal
ble approach based on extreme learning machines is adopted
in [17] to predict stability after a disturbance, while a recur-
rent neural network with long short term memory (LSTM)
cells is trained in [18] for the same task considering time
dependencies in the data. Furthermore, prediction of damping
state (i.e. well or poorly damped) is proposed by [19] using
decision trees and by [20] using neural networks.
Driven from the above, this paper presents an intelligent
Power Oscillation Damper (iPOD) that enables the SPC to
adapt for maximum -oscillatory- mode attenuation. The iPOD
is based on an ensemble AI predictor called Random Forests
(RF), which is trained to predict the real and imaginary FIGURE 1. Modified two area system.
TABLE 5. Parameters of PSS. FIGURE 16. Eigenvalues of system per control scenario.
V. CONCLUSION
The connection of energy systems through power electronic
devices is adding to the complexity of power systems by
altering the overall behaviour of power systems. Rotor angle
stability is of primary concern as the presence of power elec-
worth emphasizing that the improved damping from the SPC tronics in power systems is rising. Conventional PSS are per-
based on iPOD is achieved with only 1/7 of the synchronous forming acceptably for attenuating oscillatory modes, how-
generator apparent power. ever, only for a narrow range of operating conditions. Power
converters controlled through grid-forming techniques, e.g.
C. iPOD COMPARISON WITH PSS a SPC augmented by an iPOD as in this paper, can provide
Commonly, the PSS is used to provide damping of critical an additional source for damping power oscillations with the
modes in a system. This section compares the individual use of data and AI.
contribution of each PSS device with the corresponding con- Random Forest is a powerful machine learning model,
tribution provided by the SPC augmented by the iPOD. For which exploit the inherent diversity of combing individual
the purpose of this study, the PSS parameters are considered decisions trees trained with random subsets of the original
fixed and their values are set following the methodology dataset. In this paper, two random forests have been devel-
described in [32] and listed in Table 5. Concretely, random oped for predicting the real and imaginary component of the
points are generated following the aforementioned process inter-area mode in the two are system. These predictions are
and for each control scenario modal analysis of the system used to tune the iPOD added to SPC for improving damping
is performed. The results, averaged over all random points, of the targeted inter-area mode. As it has been proven from
are listed in Table 6. thorough analyses, the SPC with iPOD can enhance the stabil-
Clearly, PSS1 and PSS3, on average, possess the highest ity of a power system by increasing the damping of the inter-
contribution towards inter-area power oscillation damping. area mode. In the test presented in this paper, the response
However, damping provided by PSS1 and PSS3 varies signif- of recorded frequency demonstrated significant improvement
icantly between each case. Particularly, PSS1 fails to provide when SPC is augmented by the iPOD scheme. In contrast to
damping for the inter-area mode for a random operating as conventional PSS, the iPOD does not have a phase lag and
only the frequency of the oscillation is required. Therefore, [12] M. G. Taul, X. Wang, P. Davari, and F. Blaabjerg, ‘‘Current limiting
reducing the number of parameters needed to be tuned for control with enhanced dynamics of grid-forming converters during fault
conditions,’’ IEEE J. Emerg. Sel. Topics Power Electron., vol. 8, no. 2,
attenuating the inter-area mode. pp. 1062–1073, Jun. 2020.
Furthermore, the proposed scheme is shown to be robust [13] P. Rodriguez, C. Citro, J. I. Candela, J. Rocabert, and A. Luna, ‘‘Flexible
against network variations, performing equally and even grid connection and islanding of SPC-based PV power converters,’’ IEEE
Trans. Ind. Appl., vol. 54, no. 3, pp. 2690–2702, May 2018.
better than the conventional PSS in the system. However, [14] A. Tarraso, N.-B. Lai, G. N. Baltas, and P. Rodriguez, ‘‘Power quality
the proposed control scheme should not be considered as services provided by virtually synchronous FACTS,’’ Energies, vol. 12,
an adversary of PSS but rather as a complementary ally no. 17, p. 3292, 2019. [Online]. Available: https://www.mdpi.com/1996-
1073/12/17/3292
for improving the performance of the power systems of the [15] P. R. Cortes, J. I. C. Garcia, J. R. Delgador, and R. Teodorescu, ‘‘Virtual
future, which will be dominated by power electronics. The controller of electromechanical characteristics for static power convert-
increased utilization of grid-forming converters for additional ers,’’ Europen Patent 2 683 075 A1, Jan. 8, 2014.
[16] M. Pérez-Ortiz, S. Jiménez-Fernández, P. Gutiérrez, E. Alexandre,
functionalities other than their primary purpose will increase C. Hervás-Martínez, and S. Salcedo-Sanz, ‘‘A review of classification
flexibility and robustness in the ever-changing power system problems and algorithms in renewable energy applications,’’ Energies,
environment without the necessity for additional investments. vol. 9, no. 8, p. 607, 2016. [Online]. Available: https://www.mdpi.com/
1996-1073/9/8/607
[17] G. N. Baltas, C. Perales-González, P. Mazidi, F. Fernandez, and
ACKNOWLEDGMENTS P. Rodriguez, ‘‘A novel ensemble approach for solving the transient sta-
This work was supported in part by the European Com- bility classification problem,’’ in Proc. 7th Int. Conf. Renew. Energy Res.
Appl. (ICRERA), Oct. 2018, pp. 1282–1286.
mission under Project FLEXITRANSTORE—H2020-LCE- [18] J. J. Q. Yu, D. J. Hill, A. Y. S. Lam, J. Gu, and V. O. K. Li, ‘‘Intelligent
2016-2017-SGS-774407, and in part by the Spanish time-adaptive transient stability assessment system,’’ IEEE Trans. Power
Ministry of Science under Project ENE2017-88889-C2-1-R. Syst., vol. 33, no. 1, pp. 1049–1058, Jan. 2018.
[19] P. McNabb, D. Wilson, and J. Bialek, ‘‘Classification of mode damping and
Any opinions, findings and conclusions or recommendations amplitude in power systems using synchrophasor measurements and clas-
expressed in this material are those of the authors and do not sification trees,’’ IEEE Trans. Power Syst., vol. 28, no. 2, pp. 1988–1996,
necessarily reflect those of the host institutions or funders. May 2013.
[20] F. Sulla, E. Måsbäck, and O. Samuelsson, ‘‘Linking damping of elec-
tromechanical oscillations to system operating conditions using neural
REFERENCES networks,’’ in Proc. IEEE PES Innov. Smart Grid Technol., Eur., Oct. 2014,
pp. 1–6.
[1] L. Meegahapola and D. Flynn, ‘‘Impact on transient and frequency stability
[21] P. Kundur, N. J. Balu, and M. G. Lauby, Power System Stability and
for a power system at very high wind penetration,’’ in Proc. IEEE PES Gen.
Control (The Epri Power System Engineering). New York, NY, USA:
Meeting, Jul. 2010, pp. 1–8.
McGraw-Hill, 1994.
[2] High Penetration of Power Electronic Interfaced Power Sources [22] R. Badar and J. Shair, ‘‘Adaptive NeuroFuzzy sliding mode based damping
(HPoPEIPS) ENTSO-E Guidance Document for National Implementation control for SSSC,’’ in Proc. SAI Intell. Syst. Conf. (IntelliSys), Y. Bi,
for Network Codes on Grid Connection, ENTSO-E, Brussels, Belgium, S. Kapoor, and R. Bhatia, Eds. Cham, Switzerland: Springer, 2018,
2017. pp. 91–106.
[3] Z. A. Obaid, L. M. Cipcigan, and M. T. Muhssin, ‘‘Power system [23] R. O. Duda, P. E. Hart, and D. G. Stork, Pattern Classification, 2nd ed.
oscillations and control: Classifications and PSSs’ design methods: Hoboken, NJ, USA: Wiley, 2001.
A review,’’ Renew. Sustain. Energy Rev., vol. 79, pp. 839–849, Nov. 2017. [24] S. Theodoridis and K. Koutroumbas, Pattern Recognition, 4th ed.
[Online]. Available: http://www.sciencedirect.com/science/article/pii/ New York, NY, USA: Academic, 2008.
S1364032117307499 [25] L. Breiman, ‘‘Random forests,’’ Mach. Learn., vol. 45, no. 1, pp. 5–32,
[4] I. Kamwa, R. Grondin, and G. Trudel, ‘‘IEEE PSS2B versus PSS4B: Oct. 2001, doi: 10.1023/A:1010933404324.
The limits of performance of modern power system stabilizers,’’ IEEE [26] S. Shalev-Shwartz and S. Ben-David, Understanding Machine Learning:
Trans. Power Syst., vol. 20, no. 2, pp. 903–915, May 2005. From Theory to Algorithms. Cambridge, U.K.: Cambridge Univ. Press,
[5] IEEE Recommended Practice for Excitation System Models for Power 2014.
System Stability Studies, IEEE Standard 421.5-2016 (Revision IEEE Std [27] F. Pedregosa, G. Varoquaux, A. Gramfort, V. Michel, B. Thirion, O. Grisel,
421.5-2005), Aug. 2016, pp. 1–207. M. Blondel, P. Prettenhofer, R. Weiss, V. Dubourg, J. Vanderplas,
[6] F. Rashidi and M. Rashidi, ‘‘Robust and adaptive tuning of power A. Passos, D. Cournapeau, M. Brucher, M. Perrot, and E. Duchesnay,
system stabilizers using artificial neural networks,’’ in Innovations in ‘‘Scikit-learn: Machine learning in Python,’’ J. Mach. Learn. Res., vol. 12,
Applied Artificial Intelligence. Berlin, Germany: Springer, Jul. 2004, pp. 2825–2830, Nov. 2011.
pp. 1023–1032. [28] K. Diakos, Q. Wu, and A. H. Nielsen, ‘‘Phasor measurement unit and
[7] J. Fang, H. Li, Y. Tang, and F. Blaabjerg, ‘‘On the inertia of future more- phasor data concentrator test with real time digital simulator,’’ in Proc.
electronics power systems,’’ IEEE J. Emerg. Sel. Topics Power Electron., IEEE PES Asia–Pacific Power Energy Eng. Conf. (APPEEC), Dec. 2014,
vol. 7, no. 4, pp. 2130–2146, Dec. 2019. pp. 1–5.
[8] W. Zhang, A. Tarraso, J. Rocabert, A. Luna, J. I. Candela, and P. Rodriguez, [29] T. S. Ustun, S. M. Farooq, and S. M. S. Hussain, ‘‘A novel approach for
‘‘Frequency support properties of the synchronous power control for mitigation of replay and masquerade attacks in smartgrids using IEC 61850
grid-connected converters,’’ IEEE Trans. Ind. Appl., vol. 55, no. 5, standard,’’ IEEE Access, vol. 7, pp. 156044–156053, 2019.
pp. 5178–5189, Sep. 2019. [30] T. S. Ustun, S. M. Farooq, and S. M. S. Hussain, ‘‘Implementing secure
[9] J. Rocabert, A. Luna, F. Blaabjerg, and P. Rodríguez, ‘‘Control of power routable GOOSE and SV messages based on IEC 61850-90-5,’’ IEEE
converters in AC microgrids,’’ IEEE Trans. Power Electron., vol. 27, Access, vol. 8, pp. 26162–26171, 2020.
no. 11, pp. 4734–4749, Nov. 2012. [31] M. Abdollahi, J. I. Candela, J. Rocabert, M. A. Elsaharty, and P. Rodriguez,
[10] S. Yazdani, M. Ferdowsi, M. Davari, and P. Shamsi, ‘‘Advanced current- ‘‘Novel analytical method for dynamic design of renewable SSG SPC unit
limiting and power-sharing control in a PV-based grid-forming inverter to mitigate low-frequency electromechanical oscillations,’’ IEEE Trans.
under unbalanced grid conditions,’’ IEEE J. Emerg. Sel. Topics Power Power Electron., vol. 35, no. 7, pp. 7532–7544, Jul. 2020.
Electron., vol. 8, no. 2, pp. 1084–1096, Jun. 2020. [32] A. Stativă and F. Gonzalez-Longatt, ‘‘Peer-to-peer (P2P) MATLAB–
[11] A. Tayyebi, D. Groß, A. Anta, F. Kupzog, and F. Dörfler, ‘‘Frequency PowerFactory communication: Optimal placement and setting of power
stability of synchronous machines and grid-forming power converters,’’ system stabilizer,’’ in Advanced Smart Grid Functionalities Based on
IEEE J. Emerg. Sel. Topics Power Electron., vol. 8, no. 2, pp. 1004–1018, PowerFactory. Cham, Switzerland: Springer, 2018, pp. 301–318, doi:
Jun. 2020. 10.1007/978-3-319-50532-9_12.
GREGORY N. BALTAS (Student Member, IEEE) ANDRÉS TARRASÓ (Student Member, IEEE)
received the B.S. degree in electrical engineering received the M.Sc. degree in electrical engineer-
from the Technological Educational Institute of ing from the Technical University of Catalonia,
Central Greece, Chalkis, Greece, in 2015, and the Barcelona, Spain, in 2017. He is currently pursu-
M.S. degree in power engineering from the Univer- ing the Ph.D. degree with the Technical Univer-
sity of Sydney, Sydney, Australia. He is currently sity of Catalonia, Barcelona. From 2017, he has
pursuing the Ph.D. degree in data science with the been a Research Assistant with the Department
Universidad Loyola Andalucía, of Electrical Engineering, Technical University of
Since 2018, he has been a Research Assistant Catalonia, where he has been a Researcher and an
with the Loyola Institute of Science and Tech- Assistant Professor with the Department of Elec-
nology (Loyola.TECH), Universidad Loyola Andalucía. His research inter- trical Engineering. His current research interests include power electronics,
ests include applied artificial intelligence, power system stability, and photovoltaics, wind energy systems, and microgrids.
autonomous energy systems.